Flanges Research Maart 2011

8/10/2019 Flanges Research Maart 2011

1/62

anges es gn assessmenRotterdam, 10 March 2011


2/62

Scope

. n ro uc on

2.Flange failures & flange design

3.Flange qualification & load assessment methods

4.Present ISO 14692 approach and update

Copyright 2010 Dynaflow Research Group BV


3/62

FRP pipe pressure versus diameter Jointing systems

90

100

70

g]

. Cemented & Laminated(flanges)

40

50

Pres

sure[Bar , , ,

Flanges , RSJ, Double Bellcouplers

20

30Design

Double Bell Couplers & Laminations,

Flanges

0

10


Nominal Diameter [mm]


4/62


5/62

hortcomin s of resent RP codes

SIFs and flexibil ities

Local Buckl ing of large bore U/G headers (limited effect of side

support)

Interference of underground pipes

A/G supporting of large bore headers



6/62

ar ous ypes o anges an manu ac ur ng me o s

-

Transfer (compression) Molded

Flange resins: Epoxy

(Polyester)

Rigid f langes

Cemented flanges

Loose flanges



7/62



8/62


9/62



10/62

Crack in the i e ad acent to the flan e

EXAMPLE



11/62

crack in the pipe



12/62

between the bolts and

inside the bolt hole



13/62

Crack at the hub neck

Copyright 2010 Dynaflow Research Group BV 13


14/62

Crack at the hub neck



15/62

Sheared flan e startin with crack at the neck of the hub



16/62

Crack at the hub neck EXAMPLE



17/62

EXAMPLECrack at the hub neck



18/62

crac a e u nec



19/62



20/62

Intermediate conclusion

rac s at t e u nec are t e om nat ng a ure mec an sm

Cracks at the hub neck are often catastrofic



21/62

Cracks at the hub neck are related to a match-up problem between flange

r ng an connec e p pe

tM=

3

,1 **)*( pipepipeaxialpipepipe tEtDfk


flangeringpipe + ,2 **),( flangeflangeringntialcircumfereioflangering tEDDfk


22/62


23/62

Flan e Crackin is hinderin FRP a lication and has to be eleminated

ross ange a ure s n a e y crac s an s ca as rop c

Flange cracks are relatively common (more common in larger bore flanges)

A cracked flange is normal

Most cracks are superficial (only in resin rich area and not penetrating

into reinforced flange body)

Torque of the bolts

Excessive external loads (moments)

Application of wrong gasket

e ect ve ange es gn

Defective flange production



24/62

Su erficial cracks ??



25/62

Superficial cracks ??



26/62

Scope

. n ro uc on

2.Flange failures





27/62

Flange Assessment/Design methods:

Some only internal pressure

External loads incorporated as increased internal pressure (equivalent

pressure

Dedicated FRP codesCommon Assessment Items:

1.Longitudinal Stress in Hub

ASME RTRP2.Radial Stress in the flange

3.Tangential/Circumferential Stress in the flange

4.Largest combined stress (Hub + Flange)

5.Radial stress at the bolt circle

Metal (Isotropic) codes:

AD Merkblatter

RToD D-0701 Key:

EN 13480 EN 1591

UNI 2231

What is allowable stress??

Depends on location and flange manufacturing method


Based on supplier experience??

Based on tests??


28/62

Crack at the hub neck due to external moments??



29/62

Flange Failure due to external loads??EXAMPLE



30/62

Flange verification:

Conform ASME RTP-1

1. Internal pressure only

2. Allowable stress 4600 psi

(supplier experience??)



31/62



32/62



33/62



34/62

x erna ange oa s are mpor an

How to address external loads

1. Present practice incl present issue of ISO 14692

2. Iso 14692 revision (2011-2012)



35/62

GRP Flange qualification conform current issue of

ISO 14692

Flanges qualified by

.

B. Testing conform ASTM D4024 / D5421



36/62

Flange Qualif ication

Pressure performance based

Short term rupture = 4 x rated pressure

External Loads???

Equivalent pressure rule??



37/62

External loads incorporated by means of Equivalent pressure rule

ec on . u a so e.g. o


internal equivalent rating


38/62

Equivalent pressure rule

Pinternal+ Pequivalent< Prating

Pequivalent= 16*M/(*G3

)*(Koves factor Fk)+4*F/(*G2

)

Koves factor Fk:

External moment factor to account for difference in

local stress and local rotation due to an equivalent

J = Polar Moment of Inertia of flange cross-sectionI = Bending Moment of Inertia of flange cross-section1

*

*1

+

= JGFk



39/62

wo ques ons:

1.Is the equivalent pressure approach also a valid approach for GRP

2.Is the Koves factor (smoothing the effect of moment loads) valid for GRP

flanges.??



40/62


Tensile Stress

FE-Check of equivalent pressure rule


41/62

Flange failure due to local loadingEXAMPLE



42/62

Flange failure due to local

loading (fish plates)



43/62

Scope

. n ro uc on

2.Flange failures & flange design





44/62

Flange Assessment/Design methods:

Dedicated FRP codes

ASME B&PV Sect ion X

ASME RTRP

Common Assessment Items:

1.Longitudinal Stress in Hub

Metal (Isotropic) codes:

AD Merkblatter RToD D-0701

EN 13480

. a a ress n e ange

3.Tangential/Circumferential Stress in the flange

4.Largest combined stress (Hub + Flange)5.Radial stress at the bolt circ le

EN 1591

UNI 2231

Alternative method:. anges a ue o s ra n:Simplified Design Method

Flange rotation criterion: 1 deg.



45/62

or ange manu ac urer

Qualification criteria for family representative flanges

For Engineer

Generation of allowable flange load envelope



46/62

GRP Flange qualification conform current issue of

ISO 14692

A. 1000 hr Qualification tests conform ASTM D1598

B. Testing conform ASTM D4024 / D5421

Note:+ Draw back of qualification method: Internal pressure loading only

+ Simulating external loads by increased qualification pressure??


+ ow to sca e ot er anges rom test resu ts


47/62

Five points

1. Pressure rating: Flanges are rated using MPR: MPR = f2 * Pq (f2 = 0.67)

2. Flange qualification:Flanges are qualif ied to Pq by means of:

Short term cyclic loading and vacuum testing

1000 hr survival testing

3. Flange scaling Flange scaling within product sectors based on stress as per ASME RTP-1=2005. In addit ion requirement on minimum f lange rigid ity as per ASME B&PV Sect VIII div 1 appendix 4

4. Long term flange load envelope The long term design envelope is demonstrated by combined load testing on

representative flange joints

5. System design conform Equivalent pressure method and flange design

envelo e.



48/62

Flange Qualification

ems o e emons ra e :

1. Long term strength over the design life of the flange

hydrotest conditions

3. No flange damage at operating and hydrotest conditions

4. Verification of recommended maximum bolt torque in combination withgasket for flange damage.

5. Verification that gasket, flange bolt torque combination can withstand

vacuum



49/62

1.Long term strength by means of 1000 hr survival test (Arrangement A) and

.

1. In the survival test (at temperature) the flange is allowed to show

amage u no a ure e.g. ea age w n e r.

2.The leak test (at ambient) is done by 10 pressure cycles for 5 min at 1.5 * MPR

-


Leakage is considered a test failure.


50/62

Applicable 1000 hr test pressure

Based on ISO defined fixed ratios typically 2.1-2.7 x MPR



51/62



52/62


53/62

Flange scaling rules

Flange scaling rules are as per ASME RTP-1-2005

Based on dimensions of qualified (family representative) flange

representative stress values are determined that can be used in scalingthe dimensions of other flanges in the same product sector

The requirement in NM12-370 that the minimum hub thickness is 50% of

the flange thickness is dropped because of interference with the bolt

.

Similar consideration for hub length. Hub length to be based on minimum

shear length.

Additional flange rigidity check is added (rigidity index between 1.0 and 1.5

at MPR



54/62



55/62



56/62


L t fl l


57/62

Long term flange envelope

1. Long term flange envelope is demonstrated by supplier by means of

1000 hr combined load tests on a 600 NB and a 1200 NB flanged joint

2. Test to be done at temperature. Therefore if required higher ratedgaskets (but of same type) may be used.

3. Two tests: 1. At the 1000 hr qualification pressure & 2. at 25% of the

1000 hr qualification pressure

4. Test data points define the basic envelope that is scaled back to the

design envelope by the pressure ratios.

. e s ape o e enve ope or o anges are equ va en ese

shapes may be scaled to other diameters and pressure classes usingthe scaling rules.


T i l fl bi d l d t t t


58/62

Typical flange combined load test set-up


Test results


59/62

Test results


Flange design envelope (conform ISO 14692 revised)


60/62

Flange design envelope (conform ISO 14692 revised)


Default f2 values


61/62

Default f2 values.



62/62


an you

Date post:	02-Jun-2018
Category:	Documents
Upload:	lauro330
View:	214 times
Download:	0 times

Flanges Research Maart 2011

Documents